Projected sea surface temperature changes in the equatorial Pacific relative to the Warm Pool edge

Abstract

Projections of equatorial sea surface temperature from CMIP5 climate models are important for understanding possible future changes in marine habitats, local rainfall and climate processes such as El Niño Southern Oscillation. Interpreting the projected changes in the tropical Pacific is complicated by the systematic cold tongue bias and overly westward location of the Warm Pool edge at the equator in coupled models. Here an index based on the maximum zonal salinity gradient is used to differentiate the Warm Pool from the cold tongue in each of 19 CMIP5 models. Warming is then calculated relative to the dynamic edge of the Warm Pool between the second halves of the 20th and 21st Centuries from the RCP8.5 scenario to provide a bias adjusted SST projection.Based on this definition of the edge, while the Warm Pool edge is projected to warm, it is likely to remain within 10° of its present longitude. This is in stark contrast to the large projected eastward displacements of the isotherms that are usually used to define the edge. Adjusting for the edge, warming within the Warm Pool is projected to be fairly uniform with surface water freshening. Projected warming is enhanced over the cold tongue with the net effect of reducing the zonal SST gradient. In contrast, if the warming is calculated without correcting for the edge of the Warm Pool, the warming signature is dominated by the poorer performing models with an overly westward Warm Pool, resulting in enhanced warming across the equatorial Pacific. Bias adjusting realigns the warming signature and reduces the model spread of projected warming. The biased warming signature also introduces spurious meridional and zonal SST gradients. This will potentially alter the behaviour of the atmospheric convergence zones and the dynamics of ENSO which is influenced by the extent of the Warm Pool and zonal SST gradients.